Identifying the Spatial Drivers of Soil Carbon-Nitrogen Stoichiometry in Germany

The carbon to nitrogen (C: N) ratio is essential for regulating soil nutrient content balance, directly influencing crop yields and microbial activity. Spatially explicit monitoring of this ratio in temperate regions remains important to fully understand its benefits. Therefore, we sought to understand the influencing factors of the C: N and map its spatial variability at 250 m/pixel across Germany. We applied n = 1687 surface soils obtained from the 2015 Land Use and Coverage Area Frame Survey (LUCAS) database coupled with key environmental covariates, including soil, climatic, human-related, topographic, and remote sensing data. A cubist machine learning model was used to relate the C: N ratio with these environmental covariates. Our analysis revealed that pH, elevation, latitude, and silt were among the top four important covariates, accounting for 76.6 % of the total variance in the C: N ratio. The Cubist model demonstrated acceptable predictive capabilities, with a root mean square error (RMSE) of 2.55 and a relatively low bias of 0.02. Our C: N prediction map indicated that the northwestern region of Germany exhibited high C: N ratio values ranging between 15 and 24. This range suggests conducive conditions that support soil microbial activity and greater nutrient availability. Furthermore, this region has high precipitation and NDVI values, corroborating our earlier point.  Our findings emphasize the importance of soil, topography, and human activity in influencing the C: N ratio in temperate regions like Germany. Thus, understanding their roles in soil stoichiometry is crucial for developing effective land management strategies to enhance soil health and agricultural productivity.

Keywords: Soil pH, Human Footprint, Nutrient Dynamics, Cubist Model, Climate Mitigation